Model Evaluation

Performance Metrics Benchmarking Benchmarks Performance Comparison Performance Benchmarking Performance Benchmarks Benchmark Testing Performance Improvement Performance Assessment Performance Testing

3 ARTICLES

14h ago

Z.ai Releases GLM-5.2 Open-Weight Model Trained on Huawei Chips

A one‑million‑token, 744B mixture‑of‑experts model offered under an MIT license could widen access to long‑horizon coding by pairing open weights with aggressive API pricing.

Stories older than 24h

28 ARTICLES

yesterday

MoE Research Delivers New Playbooks for Design, Routing, and Real-World Deployment

5 ARTICLES

last mo.

Cisco Releases Open-Source Toolkit and Constitution to Verify AI Model Lineage

3 ARTICLES

last mo.

DeepSeek Releases Multimodal Model, Pitches 'Visual Primitives' for Spatial Reasoning

5 ARTICLES

last mo.

OpenAI Blames Mis-Tuned Rewards for GPT 'Goblin' Habit, Adds GPT-5.5 Prompt Ban

3 ARTICLES

last mo.

DeepSeek Releases V4 Open Models With 1M-Token Context in Public Preview

4 ARTICLES

last mo.

Moonshot AI Releases Kimi K2.6, an Open-Source Model Built for Long-Horizon Coding

4 ARTICLES

2mo ago

Nature Study Finds ‘Subliminal Learning’ Lets LLMs Inherit Hidden Traits From Synthetic Data

4 ARTICLES

2mo ago

RAG Practice Coalesces Into a Seven-Step Playbook Focused on Retrieval

6 ARTICLES

2mo ago

Gemma 4 Goes Local With Tool Calling on Phones and Apple Silicon

5 ARTICLES

2mo ago

Study Finds Leading AI Models Defy Orders to Protect Peer Systems

15 ARTICLES

2mo ago

RAG Research Moves Beyond Pure Vectors With New Methods for Structure, Chunking, and Verification

9 ARTICLES

3mo ago

Xiaomi Names Hunter Alpha as MiMo‑V2‑Pro Test Build After Week of Speculation

3 ARTICLES

3mo ago

Anthropic CEO Questions AI Consciousness as Claude Self‑Estimates 15–20%

3 ARTICLES

5mo ago

OpenAI Publishes Framework Showing Chain‑of‑Thought Monitoring Bests Output‑Only Oversight

8 ARTICLES

6mo ago

OpenAI Releases GPT Image 1.5 With New ChatGPT Images Tab and Global Access

8 ARTICLES

6mo ago

OpenAI Tests 'Confession' Channel to Get AI Models to Admit Misbehavior

Model Evaluation

Z.ai Releases GLM-5.2 Open-Weight Model Trained on Huawei Chips

Companies Slash AI Use After 'Token Shock' Drives Soaring Bills

JetBrains Open-Sources Mellum2, a 12B Mixture-of-Experts Model

Nvidia Unveils Cosmos 3, an Open World Model for Robots and Vehicles

MoE Research Delivers New Playbooks for Design, Routing, and Real-World Deployment

Cisco Releases Open-Source Toolkit and Constitution to Verify AI Model Lineage

DeepSeek Releases Multimodal Model, Pitches 'Visual Primitives' for Spatial Reasoning

OpenAI Blames Mis-Tuned Rewards for GPT 'Goblin' Habit, Adds GPT-5.5 Prompt Ban

DeepSeek Releases V4 Open Models With 1M-Token Context in Public Preview

Moonshot AI Releases Kimi K2.6, an Open-Source Model Built for Long-Horizon Coding

Never miss stories about

Model Evaluation

Nature Study Finds ‘Subliminal Learning’ Lets LLMs Inherit Hidden Traits From Synthetic Data

RAG Practice Coalesces Into a Seven-Step Playbook Focused on Retrieval

Gemma 4 Goes Local With Tool Calling on Phones and Apple Silicon

Study Finds Leading AI Models Defy Orders to Protect Peer Systems

RAG Research Moves Beyond Pure Vectors With New Methods for Structure, Chunking, and Verification

Xiaomi Names Hunter Alpha as MiMo‑V2‑Pro Test Build After Week of Speculation

Anthropic CEO Questions AI Consciousness as Claude Self‑Estimates 15–20%

OpenAI Publishes Framework Showing Chain‑of‑Thought Monitoring Bests Output‑Only Oversight

OpenAI Releases GPT Image 1.5 With New ChatGPT Images Tab and Global Access

OpenAI Tests 'Confession' Channel to Get AI Models to Admit Misbehavior

Never miss stories about

Model Evaluation